Originally appeared in The New Republic
Last week, seven thousand delegates traveled to Chicago for the annual convention of the National Education Association, the nation’s largest teachers' union. Most of the meeting was in keeping with the NEA’s reputation for dogmatic opposition to education reform. New Business Item #93 slammed Teach For America , which recruits graduates of top colleges to teach in high-poverty schools. Another resolution described 13 things the NEA hates about U.S. Secretary of Education Arne Duncan. One delegate running for a seat on the NEA executive committee gave a speech comparing anti-union politicians to Hitler. He was elected, defeating an incumbent, soon thereafter.
But the most consequential vote of the convention was also the biggest departure from the union’s usual hard-line stance. After months of internal negotiations, the NEA endorsed the use of student standardized test scores, along with other measures, to evaluate individual teachers. Delegates who oppose standardized testing on general principle cried “sellout,” and worse. But by then it was too late. The last major interest group opposing rational teacher evaluation had bowed to the inevitable. In doing so, it gave a boost to President Obama’s domestic agenda, clarified the future course of public education, and offered a hopeful lesson for those who believe that facts and logic can bend the arc of public policy toward the greater good.
SINCE THE RAPID unionization of the profession in the 1960s, teaching has consisted of highly-skilled, well-educated professionals stuck in locally-controlled civil service regimes, which were grafted, in turn, onto industrial unionism. The inherent contradictions of this peculiar combination ultimately led to the historic NEA vote in Chicago.
Union-dominated civil service meant that, from the outside, all teachers looked the same. Everyone had to get a bachelor’s degree, usually from a state university, and obtain a standard state teaching license before entering the classroom. Once they entered the system, everyone was paid according to the same union salary schedule, which doled out raises based on years of experience and credentials such as master’s degrees. There were few differences in rank or status, meaningful performance evaluations were non-existent, and it was nearly impossible to be fired for cause.
On the inside, however, teachers looked very different. Until the 1990s, standards and curricula were largely left in the hands of local districts and schools. Principals, in turn, gave individual teachers broad discretion over what happened every morning after students settled into their desks and the classroom door closed behind them. That meant huge variance in what students were taught, even among students with different teachers in the same school and grade.
It also meant huge variance in how well students were taught. Teaching is extremely complicated and difficult to master, particularly in America’s unusually open and pluralistic public education system. As with all high-skill professions, some people are much better at it than others. State licensing regimes didn’t change that, just as some lawyers pass the bar exam and argue in front of the Supreme Court while others take out advertisements on the back of bus stop benches.
But those differences were obscured by the seeming uniformity created by civil service and union rules. That fit with the values and ambitions of teachers' unions, which saw unity, and thus power, in sameness. United they stood, soon becoming one of the most influential forces in American politics. Local, state and federal lawmakers took their money and votes and kept the uniform system going in exchange.
Then, in the late 1980s, the educational standards-and-accountability movement took root. Spurred by fears of genius children being produced en masse by foreign competitors (Soviet and Japanese children occupied the roles now being played by the Indians and Chinese), state governments established academic standards for all children and began administering annual standardized tests to see if those goals were being met.
At the time, nobody seriously proposed using the tests to evaluate individual teachers—the unions made sure of that. But, largely for administrative convenience, some states began storing student test scores in large electronic databases. Meanwhile, Moore’s law was putting inexpensive computing power in the hands of economists with an interest in education. They analyzed hundreds of thousands of test scores, calculating how much students in a given classroom improved on standardized tests over the course of a year. The results proved what everyone intuitively already knew: Some teachers were much better than others who in all other respects looked the same.
These changes set in motion a rhetorical battle that lasted for the better part of two decades before culminating last week. Teachers' unions were caught in a contradiction. They needed to make a strong general case for the importance of teachers—otherwise, why hire more of them to reduce class sizes? Why increase their pay? At the same time, they needed to deny the variable importance of individual teachers—otherwise, why shouldn’t the best get paid more money? Why should the worst be allowed to teach at all? Once those doors opened, the whole system of unity through uniformity would be at risk.
Raw political power worked for a while. When the Bloomberg administration proposed using test scores to decide whether New York City teachers should get lifetime job security, the union ran to the state legislature  and made the plan illegal. When California Democrat George Miller—as stalwart a defender of organized labor as one can find in the United States Congress—proposed new federal policies aimed at tying teacher pay to performance, he was chastised in a public hearing  by Reg Weaver, then the president of the NEA.
But in the long run, the NEA couldn’t keep fighting on every front. Journalist Steven Brill published a long, influential New Yorker article  about the New York City teachers' union’s role in keeping alcoholics, incompetents, and malcontents on the public payroll (as did NPR  and The New Republic’ s Seyward Darby , among others). In Los Angeles, the local union opposed an ACLU lawsuit aimed at ending the school district’s practice of laying off teachers based on seniority instead of performance. When an ostensibly liberal group finds itself alienating NPR listeners, New Yorker readers, and ACLU donors, it is in a lonely place indeed.
Meanwhile, other teacher organizations had ideas of their own. The smaller but highly influential American Federation of Teachers began taking steps to find common ground with education reformers. The Denver teachers' union forged an innovative agreement  with management that used test scores, along with other factors, to set teacher pay. Just weeks before the Chicago NEA convention, Illinois Governor Pat Quinn, a Democrat, signed a sweeping, union-endorsed education reform bill  that will tie teacher pay and job security to student performance. Other states and cities are moving in the same direction. The NEA could see where the crowd was going and decided that it needed to follow so it could lead.
THE VOTE WAS a boost for the Obama administration, which has made teacher quality a major part of its education agenda. More broadly, it’s a triumph of empiricism and common sense. Academic studies really can, at times, make a difference in the public debate, and even entrenched interest groups can’t defend the indefensible forever. In these times especially, this is welcome news.
This doesn’t mean, of course, that classroom teachers will soon be hired and fired solely based on how their students perform on standardized tests. Even the most aggressive reformers are only proposing that 50 percent of an individual teacher’s evaluation be based on tests; in most cases it will be substantially less. Many teachers, moreover, don’t teach in grades or subjects where standardized tests apply. The dominant method of teacher evaluation will likely be test scores combined with some form of structured classroom observation by outside experts.
Which is as it should be. Tests don’t measure the totality of student learning, and some standardized tests are cheap and poorly designed. The Obama administration has poured hundreds of millions of dollars into building better tests, but they’re still a work in progress. Nor do tests show how teachers support other teachers or help students develop things like leadership and character. That’s why the best teacher evaluation systems combine information from multiple sources including tests, observations, and peer review. Deciding how to properly weigh and analyze these variables will be no easy task. The coming years will also require difficult decisions about how to interpret this new information when deciding who gets to be a public school teacher, how they are trained, what kinds of responsibilities they are given, and how much they should be paid.
What they won’t involve is much more principled debate about whether using test scores and other methods to evaluate teachers is a good idea in the first place. The remaining holdouts will eventually give in. And that means that future NEA conventions should probably spend less time debating the health risks of soy-based products (New Business Item #9) and more time conceiving of what a union of non-uniform education professionals ought to look like. Teachers can still speak with a collective voice while acknowledging that some are stronger than others. Indeed, sophisticated new teacher evaluation systems will provide strong evidence that some teachers are, as unions have long insisted, drastically under-appreciated and underpaid. But only some of them. The era of unity through uniformity is drawing to a close.